Aims Prenatal alcohol exposure is common with 20%–40% of mothers in the UK drinking in pregnancy.

The serious effects of alcohol on the developing fetus have been well described: with effects on growth, facial morphology and the central nervous system(CNS). Some regions of the brain appear to be disproportionately affected e.g. corpus callosum and caudate nucleus.

The primary aim of this study was to produce an objective evaluation of the morphological relationship between the face, brain structure and cognitive outcome. By measuring the face-brain morphology of the newborn infant, then it might be possible to predict future outcome at the earliest possible stage.

Methods Structural brain MRI and high-resolution 3D facial images were obtained of prenatal alcohol exposed (n=76) and unexposed control(n=45) Caucasian children (age range 6.8–17.7 years). Extending existing 3D modelling techniques allowed us to build morphometric shape models of individual regions of the face, together with the shape of the corpus callosum. These joint models enable statistical analysis of correlations between neurocognitive measures and joint face-brain shape variation. in both unexposed and alcohol exposed individuals. In the second phase of this analysis, we are acquiring 3D and 2D images of the face, and a midline representation of the corpus callosum using transfontanelle ultrasound in newborn infants.

Results Using individual analysis of facial form and brain regions in 3D, we were able to accurately identify individuals with foetal alcohol syndrome, a diagnosis at the most severe end of the alcohol-exposed spectrum. Using combined modelling of face and brain, these results improved the accuracy of discrimination (ROC >0.95). Additionally, we observed a loss of normal pattern and asymmetry in the caudate nucleus, which were correlative with neurocognitive performance. In the latter phase of this study, we designed a protocol for imaging corpus callosum, and applying statistical modelling techniques for face-brain morphological analysis in neonates.

Conclusions There is growing evidence of a correlation between central facial and central brain morphology.

Identifying facial dymorphia, structural brain anomalies and the relations between them and cognition is important as it could ultimately help determine the extent of neuro-structural damage caused by in utero alcohol exposure, allowing identification of risk groups and improving earlier diagnosis.